KR102402011B1 - Medical image process apparatus and medical image learning method, and medical image process method - Google Patents

Abstract

A method for learning a medical image of a medical image processing apparatus according to an embodiment includes preparing body X-ray images for training as an input in a training data set, and individual X-ray images for training corresponding to each of the training body X-ray images as a label in a training data set. It includes the steps of preparing the body's in-body diagnostic index information, and training the artificial neural network model using the training data set.

Description

Medical image processing apparatus and its medical image learning method and medical image processing method

The present invention relates to a method and apparatus for learning a medical image of a body, and a method and apparatus for processing a medical image of the body.

A medical imaging apparatus is a device for acquiring an image of a body internal structure of a subject to be diagnosed. The medical imaging apparatus is a non-invasive examination apparatus, and shows structural details in the body, internal tissues and fluid flow, etc. by photographing and processing the image to the user. A user such as a doctor may diagnose a patient's health condition and disease by using a medical image output from the medical imaging apparatus.

As a medical imaging device, an X-ray imaging device that irradiates X-rays to an object and detects X-rays passing through the object to image an image, and magnetic resonance imaging (MRI) to provide a magnetic resonance image : There are a magnetic resonance imaging) device, a computed tomography (CT) device, and an ultrasound diagnostic device.

Recently, due to the development of image processing technology such as machine learning, a medical imaging apparatus analyzes an acquired medical image to detect or generate an analysis result of an abnormal region, which is a region where an abnormality occurs in an object. can do. As described above, an image regenerated by analyzing the acquired medical image is called a reading aid image. When the reading aid image is provided to the doctor, the doctor may more easily diagnose whether an abnormality has occurred in the object by referring to the reading aid image.

On the other hand, chest X-ray images are being used to detect and diagnose various lung diseases. A doctor acquires in-body diagnostic indicators such as cardiovascular border from a chest X-ray image of a subject, and detects lung diseases based on this. or can be diagnosed.

However, only a highly skilled doctor can accurately obtain in vivo diagnostic indicators such as cardiovascular system boundaries from a chest X-ray image, and the accuracy of the in vivo diagnostic index obtained from a chest X-ray image depends on the doctor's skill level. There was a change problem.

Korean Patent Publication No. 10-2019-0106403 (published on September 18, 2019)

According to the embodiment, there is provided a medical image learning method for learning an artificial neural network model using body X-ray images and in-body diagnostic index information, and a medical image processing apparatus thereof.

In addition, there are provided a medical image processing method and a medical image processing apparatus in which a learned artificial neural network model acquires in-body diagnostic index information from a body X-ray image of a subject to be diagnosed.

The problems to be solved of the present invention are not limited to those mentioned above, and other problems to be solved that are not mentioned will be clearly understood by those of ordinary skill in the art to which the present invention belongs from the following description.

According to a first aspect, a method for learning a medical image of a medical image processing apparatus includes preparing body X-ray images for learning as an input in a learning data set, and corresponding to each of the learning body X-ray images as a label in the learning data set preparing the in-body diagnostic index information of the individual body, and training the artificial neural network model using the training data set.

According to the second aspect, the computer program of the computer-readable recording medium storing the computer program includes instructions for, when executed by a processor, to cause the processor to perform the method for learning a medical image of the medical image processing apparatus. .

A medical image processing method using a medical image processing apparatus according to a third aspect comprises the steps of: receiving a body X-ray image of a diagnosis subject to an artificial neural network model trained on medical images; and outputting the in-body diagnostic index information obtained through the neural network model.

According to a fourth aspect, the computer program of a computer-readable recording medium storing a computer program includes instructions for, when executed by a processor, to cause the processor to perform a medical image processing method using the medical image processing apparatus do.

A medical image processing apparatus according to a fifth aspect comprises: an input unit for receiving a learning data set in which an input is body X-ray images for learning and a label is information on an in-body diagnostic index of an individual body corresponding to the learning body X-ray images; and an artificial neural network model unit in which the artificial neural network model learns the training data set received through the input unit.

In a medical image processing apparatus according to a sixth aspect, an input unit for receiving an X-ray image of a body of a subject to be diagnosed, and an artificial neural network model from the X-ray image of a body of a subject to be diagnosed received through the input unit are provided as an internal diagnostic index of the subject to be diagnosed and an artificial neural network model unit for acquiring information, and an output unit for outputting in-body diagnostic index information of the subject to be diagnosed obtained by the artificial neural network model unit.

According to an embodiment, the artificial neural network model is trained so that the learned artificial neural network model can accurately acquire information on in-body diagnostic indexes, such as cardiovascular border, from a body X-ray image of a subject to be diagnosed.

Accordingly, accurate information on the internal diagnostic index of the subject to be diagnosed is output, the information of the internal diagnostic index is displayed and output on the body X-ray image of the subject, a numerical value based on the information of the internal diagnostic index of the subject to be diagnosed is output, or acquired through learning A result of comparing the reference value of the in vivo diagnostic index information and the in vivo diagnostic index information of the subject to be diagnosed may be output.

1 is a block diagram of a medical image processing apparatus according to an exemplary embodiment.
2 is a flowchart illustrating a method of learning a medical image by a medical image processing apparatus according to an exemplary embodiment.
3 is a flowchart illustrating a medical image processing method of a medical image processing apparatus according to an exemplary embodiment.
4 to 6 are chest X-ray images illustrating in-body diagnostic index information that the medical image processing apparatus may obtain from a body X-ray image according to an exemplary embodiment.

Advantages and features of the present invention and methods of achieving them will become apparent with reference to the embodiments described below in conjunction with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below, but may be implemented in various different forms, and only these embodiments allow the disclosure of the present invention to be complete, and common knowledge in the art to which the present invention pertains It is provided to fully inform those who have the scope of the invention, and the present invention is only defined by the scope of the claims.

Terms used in this specification will be briefly described, and the present invention will be described in detail.

The terms used in the present invention have been selected as currently widely used general terms as possible while considering the functions in the present invention, but these may vary depending on the intention or precedent of a person skilled in the art, the emergence of new technology, and the like. In addition, in a specific case, there is a term arbitrarily selected by the applicant, and in this case, the meaning will be described in detail in the description of the corresponding invention. Therefore, the term used in the present invention should be defined based on the meaning of the term and the overall content of the present invention, rather than the name of a simple term.

When a part 'includes' a certain component throughout the specification, this means that other components may be further included, rather than excluding other components, unless otherwise stated.

In addition, the term 'unit' used in the specification means software or a hardware component such as an FPGA or ASIC, and 'unit' performs certain roles. However, 'part' is not limited to software or hardware. The 'unit' may be configured to reside on an addressable storage medium or it may be configured to refresh one or more processors. Thus, as an example, 'part' refers to components such as software components, object-oriented software components, class components and task components, processes, functions, properties, procedures, subroutines, segments of program code, drivers, firmware, microcode, circuitry, data, databases, data structures, tables, arrays and variables. Functions provided within components and 'units' may be combined into a smaller number of components and 'units' or further divided into additional components and 'units'.

Hereinafter, with reference to the accompanying drawings, embodiments of the present invention will be described in detail so that those of ordinary skill in the art can easily carry out the present invention. And in order to clearly explain the present invention in the drawings, parts irrelevant to the description will be omitted.

In the present specification, a subject or patient to be diagnosed may include a human or an animal, or a part of a human or animal.

Also, in this specification, 'image' may mean multi-dimensional data composed of discrete image elements (eg, pixels in a two-dimensional image and voxels in a three-dimensional image). have.

FIG. 1 shows the configuration of a medical image processing apparatus 100 according to an exemplary embodiment, but what is shown in FIG. 1 is merely exemplary. The medical image processing apparatus 100 may be implemented in a PC or a server, or may include it.

Referring to FIG. 1 , the medical image processing apparatus 100 includes an input unit 110 and an artificial neutral network model unit 120 , and according to an embodiment, an output unit 140 and an information processing unit ( 140) or a storage unit 150 may be further included.

The input unit 110 receives a training data set for learning the artificial neural network model of the artificial neural network model unit 120 . In addition, the input unit 110 receives the body X-ray image of the subject to be diagnosed to be input to the learned artificial neural network model. For example, body X-ray images for learning may be input as an input of the learning data set, and internal diagnostic index information of an individual body corresponding to each of the learning body X-ray images may be input as a label of the learning data set. For example, the in-body diagnostic index information may include information about a cardiovascular border. For example, information on the cardiovascular boundary includes the aortic knob, the lung cone (PC), the left atrial appendage (LAA), and the right upper cardiac boundary (Cardiac Border, CB), right lower cardiac border, left lower cardiac border, descending aorta (DAO), carinal, upper diaphragm, left pulmonary vein, or posterior margin of the heart ( It may include a posterior border, an anterior border, and the like.

The artificial neural network model unit 120 includes an artificial neural network model for learning the training data set received through the input unit 110 . The training data set that the artificial neural network model learns may be body X-ray images for learning, and labels may be body diagnostic index information of an individual body corresponding to each of the body X-ray images for training. In addition, the learned artificial neural network model of the artificial neural network model unit 120 may acquire the body diagnostic index information of the subject to be diagnosed from the body X-ray image of the subject to be diagnosed received from the input unit 110 . For example, the artificial neural network model may be a Mask R-CNN model. The Mask R-CNN model can use body X-ray images for training as training images of the Mask R-CNN model. (Mask) is available. For example, when the label of the training data set input through the input unit 110 includes information on the cardiovascular system boundary line as the in-body diagnostic index information of the individual body corresponding to each of the body X-ray images for training, the artificial neural network model The trained artificial neural network model of the unit 120 is the body diagnostic index information of the subject to be diagnosed, the chest length (TL), the distance between the right axis and the left axis of the heart, and the cardiothoracic ratio. , CTR), curvatures and lengths of the aortic eminence, curvature and length of the lung cone, curvature and length of the left atrial appendage, curvature and length of the right right cardiac border, the curvature and length of the right lower cardiac border, the left lower cardiac border The curvature and length of the descending aorta, the angle of the bifurcation, the cardiac area, the relative positions of the diaphragm and the ribs, the mean diameter of the right pulmonary artery, or the distance between the posterior and anterior vertebrae of the heart -to-spine) may include information about at least one of the average diameter.

In addition, the artificial neural network model unit 120 may acquire a reference value for in-body diagnostic index information of body X-ray images for learning according to the learning result using the learning data set, and compare the obtained reference value and the in-body diagnostic index information of the subject to be diagnosed. Comparison results can be drawn. For example, the artificial neural network model unit 120 may include a memory for storing instructions programmed to perform a function as an artificial neural network model, and a microprocessor for executing these instructions.

The output unit 130 may output the internal diagnostic index information of the subject to be diagnosed obtained by the artificial neural network model unit 120 to the outside. For example, the output unit 130 may output in the form of marking the internal diagnostic index information on the body X-ray image of the subject to be diagnosed, and the form of position information on the body X-ray image of the body diagnostic index information of the subject to be diagnosed ( Example: position coordinates) can also be output. In addition, the output unit 130 may output a result of comparing the reference value of the body diagnostic index information of the body X-ray images for learning derived by the artificial neural network model unit 120 with the body diagnostic index information of the subject to be diagnosed to the outside. have. In addition, when the information obtained by the artificial neural network model unit 120 is processed and provided in a predetermined form by the information processing unit 140, the output unit 130 may output the processed information in a recognizable form externally. have. The output unit 130 may include a port for outputting in-body diagnostic index information, a wired communication module, or a wireless communication module. Alternatively, the output unit 130 may include an image display device capable of outputting the diagnosis target's internal diagnostic index information in the form of an image.

The information processing unit 140 may process various types of information obtained by the artificial neural network model unit 120 into a predetermined form and provide it to the output unit 130 . For example, the information processing unit 140 may provide the output unit 130 by processing the body diagnostic index information of the subject to be diagnosed on the X-ray image of the subject input through the input unit 110 in a marked form. have. Alternatively, the information processing unit 140 may provide the output unit 130 with a numerical value based on the in-body diagnostic index information of the subject to be diagnosed obtained by the artificial neural network model unit 120 . Alternatively, the information processing unit 140 compares the reference value of the body diagnostic index information of the body X-ray images for learning derived by the artificial neural network model unit 120 with the body diagnostic index information of the subject to be diagnosed in a predetermined form. It may be processed and provided to the output unit 130 . For example, the information processing unit 140 may include a memory for storing instructions programmed to perform a function of processing various types of information into a predetermined form and a microprocessor for executing these instructions.

The storage unit 150 stores commands programmed to cause the medical image processing apparatus 100 to perform various functions, stores various information such as images received through the input unit 110 , or the artificial neural network model unit 120 . And/or the calculation and processing results by the information processing unit 140 may be stored. For example, the storage unit 150 includes magnetic media such as hard disks, floppy disks and magnetic tapes, optical media such as CD-ROMs and DVDs, and floppy disks. It may be a computer-readable recording medium, such as a hardware device specially configured to store and execute program instructions, such as a magneto-optical medium or a flash memory.

FIG. 2 is a flowchart illustrating a method for learning a medical image by the medical image processing apparatus 100 according to an embodiment, and FIG. 3 is a flowchart for explaining a method for processing a medical image by the medical image processing apparatus 100 according to an embodiment is a flow chart for 4 to 6 are chest X-ray images illustrating in-body diagnostic index information that the medical image processing apparatus 100 may obtain from a body X-ray image according to an embodiment.

Hereinafter, a medical image learning method and a medical image processing method of the medical image processing apparatus 100 according to an embodiment will be described in detail with reference to FIGS. 1 to 6 .

First, a medical image learning method of the medical image processing apparatus 100 will be described according to the flow sequence of FIG. 2 .

A training data set for learning the artificial neural network model of the artificial neural network model unit 120 of the medical image processing apparatus 100 is prepared. For example, as an input of the learning data set, body X-ray images for learning may be prepared ( S210 ), and as a label of the learning data set, information on individual body diagnostic indexes corresponding to each of the learning body X-ray images is provided. It can be prepared (S220).

According to an exemplary embodiment, the medical image processing apparatus 100 may use information on the cardiovascular system boundary line as shown in FIGS. 4 to 6 for in-body diagnostic index information. Cardiovascular boundary lines corresponding to reference numerals in FIGS. 4 to 6 are as follows.

403: aortic eminence

404: lung cone

405: left atrium

406: right heart borderline

407: lower right heart border

408: left lower heart borderline

409: descending aorta

410: branch

502: the top point of the diaphragm 501

504: right pulmonary artery

601: posterior edge of the heart and anterior edge of the spine

Based on the information on the cardiovascular system boundaries of reference numerals 401 to 601, the medical image processing apparatus 100 provides information on the in-body diagnostic index, including the chest length 401, the distance between the right axis of the heart and the left axis of the heart 402, cardiopulmonary coefficient, and the aorta. Curvature and length of ridge 403 , curvature and length of lung cone 404 , curvature and length of left atrial appendage 405 , curvature and length of right right cardiac borderline 406 , and curvature of right lower cardiac borderline 407 . and length, the curvature and length of the lower left cardiac borderline 408, the curvature and length of the descending aorta 409, the angle of the bifurcation 410, the heart area, the relative position of the diaphragm 501 and the ribs 503, the right The average diameter of the pulmonary artery 504 or the average diameter between the trailing edge of the heart and the anterior edge of the spine 601 may be used. Here, the relative positions of the diaphragm 501 and the ribs 503 may be used for determining whether the subject to be diagnosed has sufficiently breathed when taking an X-ray image. For example, when the upper end of the diaphragm is located below the 10th rib, it can be determined that the subject has sufficiently breathed. When an X-ray image taken in a state of not breathing enough is included in the body X-ray images for initial learning, it may be excluded from the learning data set used for actual learning so as not to be reflected in the in-body diagnostic index.

The training data set prepared through steps S210 and S220 is input through the input unit 110 and provided to the artificial neural network model unit 120, and the artificial neural network model of the artificial neural network model unit 120 is trained in steps S210 and S220. train the data set. For example, the Mask R-CNN model may be used as the artificial neural network model. In this case, body X-ray images for learning are used as training images of the Mask R-CNN model, and the internal diagnostic index information of each body corresponding to each of the learning body X-ray images is used as a mask of the Mask R-CNN model. It can be used as (S230).

Next, a medical image processing method of the medical image processing apparatus 100 will be described according to the flow sequence of FIG. 3 .

First, as described with reference to FIG. 2 , in the medical image processing apparatus 100 , the artificial neural network model of the artificial neural network model unit 120 receives body X-ray images for learning as input data and body X-ray images for learning as labels. A learning data set including information on the in-body diagnostic index of an individual body corresponding to each is learned ( S310 ).

In the state in which the artificial neural network model of the artificial neural network model unit 120 has been learned, the body X-ray image of the subject to be diagnosed is inputted through the input unit 110 and provided to the artificial neural network model unit 120 ( S320 ).

In addition, the artificial neural network model of the artificial neural network model unit 120 acquires the body diagnostic index information of the diagnosis subject from the body X-ray image of the diagnosis subject provided through step S320. Here, the in-body diagnostic index information obtained by the artificial neural network model includes chest length 401, the distance between the right and left axes of the heart 402, cardiopulmonary coefficient, the curvature and length of the aortic ridge 403, and the length of the lung cone 404. curvature and length, the curvature and length of the left atrial appendage 405, the curvature and length of the right right cardiac borderline 406, the curvature and length of the right lower cardiac borderline 407, the curvature and length of the left lower cardiac borderline 408; The curvature and length of the descending aorta 409, the angle of the bifurcation 410, the heart area, the relative positions of the diaphragm 501 and the ribs 503, the average diameter of the right pulmonary artery 504, or between the posterior margin of the heart and the anterior vertebrae ( 601) may include an average diameter (S330).

Then, the output unit 130 may output the internal diagnostic index information of the subject to be diagnosed obtained by the artificial neural network model unit 120 to the outside.

Here, the output unit 130 may output the internal diagnostic index information obtained by the artificial neural network model unit 120 in the form of raw data, but may also output it in a predetermined form processed by the information processing unit 140 . have. To this end, the information processing unit 140 may process various types of information obtained by the artificial neural network model unit 120 into a predetermined form and provide it to the output unit 130 . For example, the information processing unit 140 may provide the output unit 130 by processing the body diagnostic index information of the subject to be diagnosed on the X-ray image of the subject input through the input unit 110 in a marked form. have. Alternatively, the information processing unit 140 may provide the output unit 130 with a numerical value based on the in-body diagnostic index information of the subject to be diagnosed obtained by the artificial neural network model unit 120 . Alternatively, the information processing unit 140 compares the reference value of the body diagnostic index information of the body X-ray images for learning derived by the artificial neural network model unit 120 with the body diagnostic index information of the subject to be diagnosed in a predetermined form. It may be processed and provided to the output unit 130 . For example, information such as how many percent (%) the body X-ray image of the subject to be diagnosed is different from the reference value obtained through prior learning may be provided through step S320.

Accordingly, the output unit 130 may output in the form of marking the internal diagnostic index information on the body X-ray image of the subject to be diagnosed, and the form of location information on the body X-ray image of the body diagnostic index information of the subject to be diagnosed (eg: position coordinates) can also be output. In addition, the output unit 130 may output a result of comparing the reference value of the body diagnostic index information of the body X-ray images for learning derived by the artificial neural network model unit 120 with the body diagnostic index information of the subject to be diagnosed to the outside. have. For example, the output unit 130 may output information acquired by the artificial neural network model unit 120 recognizable from the outside ( S340 ).

Meanwhile, each step included in the medical image learning method and the medical image processing method according to the above-described embodiment may be implemented in a computer-readable recording medium for recording a computer program programmed to perform these steps.

In addition, each step included in the medical image learning method and the medical image processing method according to the above-described embodiment may be implemented in the form of a computer program stored in a computer-readable recording medium programmed to perform these steps. .

Combinations of each step in each flowchart attached to the present invention may be performed by computer program instructions. These computer program instructions may be embodied in a processor of a general purpose computer, special purpose computer, or other programmable data processing equipment, such that the instructions performed by the processor of the computer or other programmable data processing equipment provide the functions described in each step of the flowchart. It creates a means to do these things. These computer program instructions may also be stored in a computer-usable or computer-readable medium that may direct a computer or other programmable data processing equipment to implement a function in a particular manner, and thus the computer-usable or computer-readable medium. The instructions stored in the recording medium are also possible to produce an article of manufacture including instruction means for performing the functions described in each step of the flowchart. The computer program instructions may also be mounted on a computer or other programmable data processing equipment, such that a series of operational steps are performed on the computer or other programmable data processing equipment to create a computer-executed process to create a computer or other programmable data processing equipment. It is also possible that instructions for performing the processing equipment provide steps for performing the functions described in each step of the flowchart.

Further, each step may represent a module, segment, or portion of code comprising one or more executable instructions for executing the specified logical function(s). It should also be noted that in some alternative embodiments it is also possible for the functions recited in the steps to occur out of order. For example, it is possible that two steps shown one after another may in fact be performed substantially simultaneously, or that the steps may sometimes be performed in the reverse order according to the corresponding function.

The above description is merely illustrative of the technical spirit of the present invention, and various modifications and variations will be possible without departing from the essential quality of the present invention by those skilled in the art to which the present invention pertains. Accordingly, the embodiments disclosed in the present invention are not intended to limit the technical spirit of the present invention, but to explain, and the scope of the technical spirit of the present invention is not limited by these embodiments. The protection scope of the present invention should be interpreted by the following claims, and all technical ideas within the scope equivalent thereto should be interpreted as being included in the scope of the present invention.

100: medical image processing device
110: input unit
120: artificial neural network model unit
130: output unit
140: information processing unit
150: storage

Claims (15)

A medical image learning method of a medical image processing apparatus, comprising:
receiving, by the medical image processing apparatus, body X-ray images for learning as an input from among a set of learning data;
receiving, by the medical image processing device, information on individual body diagnostic indexes corresponding to each of the body X-ray images for learning as a label in the learning data set;
Including the step of learning the training data set by the artificial neural network model of the medical image processing device,
The in-body diagnostic index information includes information on the relative positions of the diaphragm and the ribs,
Among the training data sets actually used for learning the artificial neural network model, body X-ray images for training are partially excluded based on information on the relative positions of the diaphragm and ribs among the body X-ray images for initial training
How to learn medical imaging. The method of claim 1,
The in-body diagnostic index information includes information on a cardiovascular system border (Cardiovascular Border)
How to learn medical imaging. 3. The method of claim 2,
Information on the cardiovascular system boundary line, aortic knob, lung cone (Pulmonary Conus, PC), left atrial appendage (Left Atrial Appendage, LAA), right upper (right upper) heart boundary line (Cardiac Border, CB), right lower heart border, lower left heart border, descending aorta (DAO), carinal, upper diaphragm, left pulmonary vein, or posterior border , comprising at least one information of the anterior border of the spine
How to learn medical imaging. The method of claim 1,
The artificial neural network model is a Mask R-CNN model, the body X-ray images for learning are used as training images of the Mask R-CNN model, and the in-body diagnostic index information is used as a mask of the Mask R-CNN model.
How to learn medical imaging. As a computer-readable recording medium storing a computer program,
The computer program, when executed by a processor,
A computer-readable recording medium comprising instructions for causing the processor to perform the method according to any one of claims 1 to 4. A medical image processing method using a medical image processing apparatus, comprising:
receiving an X-ray image of the body of a diagnosis subject to an artificial neural network model that has learned medical images;
and outputting in-body diagnostic index information obtained through the artificial neural network model from the body X-ray image of the subject to be diagnosed,
In the artificial neural network model, the input is body X-ray images for learning, and the label includes in-body diagnostic index information of an individual body corresponding to the training body X-ray images, including information about the cardiovascular border. The data set is trained,
The body diagnostic index information of the subject to be diagnosed includes a chest length (TL), a distance between a right axis and a left axis of the heart, a curvature and length of an aortic knob, and a lung Curvature and length of the Pulmonary Conus (PC), the curvature and length of the Left Atrial Appendage (LAA), the curvature and length of the right upper heart line, the curvature and length of the right lower heart line, and Length, curvature and length of the lower left cardiac border (CB), the curvature and length of the descending aorta (DAO), angle of the carinal, cardiac area, left pulmonary vein ) or at least one of the mean diameter of the posterior edge of the heart and the mean diameter of the heart-to-spine.
Medical image processing method. delete 7. The method of claim 6,
The step of outputting the in vivo diagnostic index information may include displaying and outputting the in vivo diagnostic index information of the diagnostic subject on the body X-ray image of the diagnosis subject.
Medical image processing method. 7. The method of claim 6,
The outputting of the in-body diagnostic index information may include outputting a numerical value based on the in-body diagnostic index information of the subject to be diagnosed.
Medical image processing method. delete delete 7. The method of claim 6,
In the step of outputting the in-body diagnostic index information, the reference value for the in-body diagnostic index information of the body X-ray images for learning obtained according to the learning result of the artificial neural network model and the in-body diagnostic index information of the subject to be diagnosed are compared. to output
Medical image processing method. As a computer-readable recording medium storing a computer program,
The computer program, when executed by a processor,
13. A computer-readable recording medium comprising instructions for causing the processor to perform the method according to any one of claims 6, 8, 9 and 12. an input unit for receiving a learning data set in which the input is body X-ray images for learning and the label is information on an in-body diagnostic index of an individual body corresponding to each of the learning body X-ray images;
and an artificial neural network model unit in which an artificial neural network model learns the learning data set input through the input unit,
The in-body diagnostic index information includes information on the relative positions of the diaphragm and the ribs,
Among the training data sets actually used for learning the artificial neural network model, body X-ray images for training are partially excluded based on information on the relative positions of the diaphragm and ribs among the body X-ray images for initial training
Medical image processing device. an input unit for receiving an X-ray image of the body of the subject to be diagnosed;
an artificial neural network model unit in which the artificial neural network model acquires the body diagnostic index information of the diagnosis subject from the body X-ray image of the diagnosis subject received through the input unit;
And an output unit for outputting the in-body diagnostic index information of the subject to be diagnosed obtained by the artificial neural network model unit,
In the artificial neural network model, the input is body X-ray images for learning, and the label includes in-body diagnostic index information of an individual body corresponding to the training body X-ray images, including information about the cardiovascular border. The data set is trained,
The body diagnostic index information of the subject to be diagnosed includes a chest length (TL), a distance between a right axis and a left axis of the heart, a curvature and length of an aortic knob, and a lung Curvature and length of the Pulmonary Conus (PC), the curvature and length of the Left Atrial Appendage (LAA), the curvature and length of the right upper heart line, the curvature and length of the right lower heart line, and Length, curvature and length of the lower left cardiac border (CB), the curvature and length of the descending aorta (DAO), angle of the carinal, mean distance of the left pulmonary vein or at least one of a mean diameter between the posterior edge of the heart and the heart-to-spine.
Medical image processing device.

Images